HAL Id: hal-01991022
https://hal.sorbonne-universite.fr/hal-01991022
Submitted on 23 Jan 2019HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
weeks’ gestation: perinatal and 2-year outcomes within
a national population-based study (EPIPAGE-2)
Elsa Lorthe, Heloise Torchin, Pierre Delorme, Pierre-Yves Ancel, Laetitia
Marchand-Martin, Laurence Foix-l’Hélias, Valérie Benhammou, Catherine
Gire, Claude d’Ercole, Norbert Winer, et al.
To cite this version:
Elsa Lorthe, Heloise Torchin, Pierre Delorme, Pierre-Yves Ancel, Laetitia Marchand-Martin, et al.. Preterm premature rupture of membranes at 22–25 weeks’ gestation: perinatal and 2-year outcomes within a national population-based study (EPIPAGE-2). American Journal of Obstetrics and Gyne-cology, Elsevier, 2018, 219 (3), pp.298.e1-298.e14. �10.1016/j.ajog.2018.05.029�. �hal-01991022�
1
Preterm premature rupture of membranes at 225 weeks’ gestation: perinatal and
2-1
year outcomes within a national population-based study (EPIPAGE-2)
2
Elsa LORTHE1,2, RM, PhD, Héloïse TORCHIN1,3, MD, MSc, Pierre DELORME1,4, MD,
3
MSc, Pierre-Yves ANCEL1,5, MD, PhD, Ms Laetitia MARCHAND-MARTIN1, MSc,
4
Laurence FOIX-L’HELIAS1,2,6, MD, PhD, Valérie BENHAMMOU1, PhD, Catherine GIRE7,
5
MD, Claude d’ERCOLE8, MD, Norbert WINER9, MD, PhD, Loïc SENTILHES10, MD, PhD,
6
Damien SUBTIL11, MD, PhD, François GOFFINET1,4, MD, PhD, Gilles KAYEM1,2,12, MD,
7
PhD
8
1 Inserm UMR 1153, Obstetrical, Perinatal and Pediatric Epidemiology Research Team
9
(Epopé), Center for Epidemiology and Statistics Sorbonne Paris Cité, Département
Hospitalo-10
Universitaire Risks in pregnancy, Paris Descartes University
11
2 Sorbonne Universités, Université Pierre and Marie Curie Paris 06, Institut de Formation
12
Doctorale, 4 Place Jussieu, 75252 PARIS cedex 05, Paris, France
13
3 Neonatal Medicine and Resuscitation Service in Port-Royal, Cochin Hospital, Assistance
14
Publique-Hôpitaux de Paris, Paris, France
15
4 Department of Obstetrics and Gynecology, Cochin, Broca, Hôtel Dieu Hospital, Assistance
16
Publique-Hôpitaux de Paris, Paris, France
17
5 Unité de Recherche Clinique – Centre d’Investigations Cliniques P1419, Département
18
Hospitalo-Universitaire Risks in Pregnancy, Cochin Hotel-Dieu Hospital, Assistance
19
Publique-Hôpitaux de Paris, Paris F-75014, France
20
6 Department of Neonatalogy, Trousseau Hospital, Assistance Publique-Hôpitaux de Paris,
21
Paris, France
22
7 Department of Neonatology, North Hospital, Marseille, France
23
8 Department of Obstetrics and Gynecology, Nord Hospital, Assistance Publique des
24
Hôpitaux de Marseille, Aix Marseille Université, AMU, Marseille, France
25
9 Department of Obstetrics and Gynecology, CIC Mère enfant, University Hospital, INRA,
26
UMR 1280 Physiologie des adaptations nutritionnelles,Nantes, France
27
10 Department of Obstetrics and Gynecology, Bordeaux University Hospital, Bordeaux,
28
France
29
11 Department of Obstetrics and Gynecology, Jeanne de Flandre Hospital, Lille, France, EA
30
2694, PRES University of Lille Nord de France, 59000 Lille, France
31
12 Department of Obstetrics and Gynecology, Trousseau Hospital, Assistance
Publique-32
Hôpitaux de Paris, Paris, France
33 34
Funding sources: EPIPAGE-2 was funded by the French Institute of Public Health
35
Research/Institute of Public Health and its partners: the French Health Ministry, the National
36
Institute of Health and Medical Research (INSERM), the National Institute of Cancer, and the
37
National Solidarity Fund for Autonomy (CNSA); the National Research Agency through the
38
French EQUIPEX program of investments for the future (reference ANR-11-EQPX-0038);
39
and the PREMUP Foundation. The funders had no role in the study design, data collection
40
and analysis, decision to publish, or preparation of the manuscript.
41
Disclosure of interests: Claude d’Ercole is a consultant for the Ferring laboratory, this
42
laboratory was not involved in this work in any way. The other authors report no conflict of
43
interest.
44
Corresponding author: Dr Elsa Lorthe, INSERM U1153, Bâtiment Recherche, Hôpital
45
Tenon, 4 rue de la Chine, 75020 Paris, France. Phone +33 1 56 01 83 67, fax +33 1 56 01 71
46
88. Email: elsa.lorthe@gmail.com 47
Word count: Abstract (398 words), Text (3313 words)
2
Condensation:
49
PPROM at 22-25 weeks is associated with high incidence of mortality and morbidity, with
50
wide variations by GA at PPROM.
51
Implications and contributions:
52
A. To provide reliable and relevant data related to the prognosis of PPROM at 22-25 weeks to
53
adequately counsel parents during pregnancy and to reflect on our policies of care.
54
B. Nearly half of the fetuses are delivered within the first week. PPROM at 22-25 weeks is
55
associated with high incidence of perinatal mortality and morbidity, with wide variations by
56
gestational age at PPROM. However, a non-negligible proportion of children survive without
57
severe morbidity both at discharge and at 2 years.
58
C. This study is the first to describe and quantify perinatal and 2-years outcomes of singletons
59
and twins born after periviable PPROM, using data from a national prospective
population-60
based cohort. The use of different inception points to report rates of survival is helpful in
61
adapting information provided to parents when the GA of birth is not yet known.
62
Short title: Outcomes of pregnancies with periviable PROM
63 64 65 66 67 68 69 70
3
Abstract
71
Background: Most clinical guidelines state that with early preterm premature rupture of
72
membranes, obstetric and pediatric teams must share a realistic and individualized appraisal
73
of neonatal outcomes with parents and consider their wishes for all decisions. However, we
74
currently lack reliable and relevant data, according to gestational age at rupture of
75
membranes, to adequately counsel parents during pregnancy and to reflect on our policies of
76
care at these extreme gestational ages.
77
Objective: To describe both perinatal and 2-year outcomes of preterm infants born after
78
preterm premature rupture of membranes at 22-25 weeks’ gestation.
79
Study design: EPIPAGE-2 is a French national prospective population-based cohort of
80
preterm infants born in 546 maternity units in 2011. Inclusion criteria in this analysis were
81
women diagnosed with preterm premature rupture of membranes at 22-25 weeks’ gestation
82
and singleton or twin gestations with fetus(es) alive at rupture of membranes. Latency
83
duration, antenatal management, and outcomes (survival at discharge, survival at discharge
84
without severe morbidity, and survival at 2 years’ corrected age without cerebral palsy) were
85
described and compared by gestational age at preterm premature rupture of membranes.
86
Results: Among the 1435 women with a diagnosis of preterm premature rupture of
87
membranes, 379 were at 22-25 weeks’ gestation, with 427 fetuses (331 singletons and 96
88
twins). Median GA at preterm premature rupture of membranes and at birth were 24
89
(interquartile range 23-25) and 25 (24-27) weeks, respectively. For each gestational age at
90
preterm premature rupture of membranes, nearly half of the fetuses were born within the week
91
after the rupture of membranes. Among the 427 fetuses, 51.7% were survivors at discharge
92
(14.1%, 39.5%, 66.8% and 75.8% with preterm premature rupture of membranes at 22, 23, 24
93
and 25 weeks, respectively), 38.8% were survivors at discharge without severe morbidity and
4
46.4% were survivors at 2 years without cerebral palsy, with wide variations by gestational
95
age at preterm premature rupture of membranes. Survival at 2 years without cerebral palsy
96
was low with preterm premature rupture of membranes at 22 and 23 weeks but reached
97
approximately 60% and 70% with preterm premature rupture of membranes at 24 and 25
98
weeks.
99
Conclusion: Preterm premature rupture of membranes at 22-25 weeks is associated with high
100
incidence of mortality and morbidity, with wide variations by gestational age at preterm
101
premature rupture of membranes. However, a non-negligible proportion of children survive
102
without severe morbidity both at discharge and at 2-years’ corrected age.
103
Key words: cerebral palsy, EPIPAGE-2, preterm premature rupture of membranes, perinatal
104
outcome, periviable rupture of membranes, prematurity
105 106 107 108 109 110 111 112 113 114 115 116
5
Introduction
117
Early preterm premature rupture of membranes (PPROM), defined as PPROM at
22-118
25 weeks’ gestation, occurs in less than 1% of pregnancies and is associated with a high rate
119
of perinatal morbidity and mortality.1–4 Fetuses exposed to early PPROM face increased risks
120
of obstetric (placental abruption, cord prolapse, infection) and fetal complications (pulmonary
121
hypoplasia, limb deformities, prematurity and in utero demise)1,3,4 with short- and long-term
122
potential adverse consequences.
123
With these high risks of extreme prematurity and severe disability, antenatal care
124
requires considering the uncertainty about neonatal prognosis and the risks of severe maternal
125
complications, particularly sepsis. Management options are induction of labor, either
126
immediately3 or in cases of severe oligohydramnios or chorioamnionitis,5 or expectant
127
management with antibiotics and with steroids once viability is reached.3 Most clinical
128
guidelines state that with early PPROM, obstetric and pediatric teams must share a realistic
129
and individualized appraisal of neonatal outcomes with parents and consider their wishes for
130
all decisions.2,3,5 However, we currently lack reliable and relevant data, according to
131
gestational age (GA) at PPROM, to adequately counsel parents during pregnancy and to
132
reflect on our policies of care at these extreme GAs. Indeed, evidence-based data concerning
133
periviable complications of pregnancy are scarce: available data are mostly from small-sized
134
retrospective studies, often restricted to women eligible for expectant management, which
135
thus leads to overestimating neonatal survival.2,3,6
136
We aimed to describe and quantify both perinatal and 2-year outcomes of preterm
137
infants born after PPROM at 22 to 25 weeks’ gestation, within a prospective population-based
138
cohort at a national level.
139
6
Materials and methods
141
Setting and data collection of the EPIPAGE-2 cohort study 142
This a secondary analysis of EPIPAGE-2 (Etude épidémiologique sur les petits âges
143
gestationnels 2), a prospective, national, population-based cohort study of preterm infants
144
born in France in 2011.7 All live births, stillbirths and terminations of pregnancy at 220/7 to
145
346/7 weeks’ gestation (n=7804), whose parents had not declined to participate, were included
146
in 25 French regions involving 546 maternity units. Only one region, accounting for 2% of all
147
births in France, did not participate. The overall participation rate was 93%. The recruitment
148
periods differed by GA at birth: 22 to 26 weeks (8 months), 27 to 31 weeks (6 months) and 32
149
to 34 weeks (5 weeks). Extremely preterm births (22-26 weeks) were recruited during a
150
longer period because of their very low incidence and only a sample of moderate preterm
151
births (32-34 weeks) was recruited. Maternal, obstetric, and neonatal data were collected from
152
medical records following a standardized protocol. Full details of the cohort recruitment and
153
data collection are reported elsewhere.7 The EPIPAGE-2 cohort study was implemented to
154
describe short- and long-term outcomes among preterm infants. For that purpose, in children
155
included in follow-up, a detailed neurological and sensory examination was performed by the
156
referring physician at 2 years’ corrected age.8
157
Ethics 158
As required by French law and regulations, EPIPAGE-2 was approved by the National Data
159
Protection Authority (CNIL n°911009), the appropriate ethics committees (Consultative
160
Committee on the Treatment of Data on Personal Health for Research Purposes, reference
161
n°10.626) and the Committee for the Protection of People Participating in Biomedical
162
Research (reference CPP SC-2873).
163
Participants 164
7
Our study population included all women diagnosed with PPROM at 22 to 25 completed
165
weeks’ gestation and fetuses alive at the time of PPROM. PPROM was defined as
166
spontaneous rupture of membranes occurring at least 12 hr before birth. As recommended, the
167
diagnosis was made by the attending obstetric staff based on maternal history and sterile
168
speculum examination visualizing amniotic fluid leakage from the cervical os, with a
169
diagnostic test if necessary.3,5 Exclusion criteria were lethal malformations, triplets and
170
quadruplets (to obtain a more homogeneous population), as well as multiple pregnancies with
171
twin-to-twin transfusion syndrome (that can be responsible for both iatrogenic PPROM
172
related to fetoscopic selective laser photocoagulation and poorer neonatal outcomes). Differed
173
births or with one of the babies ineligible for analysis were also excluded.
174
French guidelines and practices 175
Overall, recommended antenatal care of women with PPROM include expectant management,
176
with antibiotics, corticosteroids from viability to 34 weeks’ gestation and, if necessary,
177
tocolysis and in utero transfer.5 Magnesium sulfate was not routinely used for tocolysis or
178
neuroprotection in 2011. According to French legislation, termination of pregnancy (TOP) on
179
parental request can be provided at any time if the fetus is affected by a severe and incurable
180
pathology or if maternal life is seriously jeopardized. With PPROM before 24 weeks’
181
gestation, guidelines from the National College of French Gynecologists and Obstetricians
182
state that medical TOP should not be considered in the absence of oligohydramnios or
183
chorioamnionitis and that all decisions should take into account parental wishes after adequate
184
counseling.5
185
Assessment of the natural history of PPROM 186
The natural history of periviable PPROM was investigated by the latency period (the time
187
elapsed from rupture to delivery), GA at birth, determined as the best obstetrical estimate
8
combining last menstrual period and first-trimester ultrasonography assessment, and the
189
specific complications of early PPROM. We focused on the following complications: severe
190
oligohydramnios in the last measurement before delivery (i.e., largest vertical pocket < 2 cm
191
or amniotic fluid index < 5, with anhydramnios defined as amniotic fluid index = 0), placental
192
abruption, cord prolapse, fetal consequences of prolonged oligohydramnios (i.e., pulmonary
193
hypoplasia and/or limb deformities) and clinical chorioamnionitis. The diagnosis of clinical
194
chorioamnionitis was not standardized in this observational cohort, but all relevant data were
195
collected and allowed us to define clinical chorioamnionitis as maternal temperature ≥ 37.8°C
196
(100°F) associated with any two of the following criteria: uterine tenderness, purulent or
foul-197
smelling amniotic fluid, maternal tachycardia, fetal tachycardia, and maternal leukocytosis ≥
198
15,000 cells/mm3. Data to assess maternal outcomes, including infectious complications, were
199
not exhaustive in the EPIPAGE 2 questionnaires and were thus not analyzed.
200
Antenatal management 201
We described antenatal care provided to women in terms of in utero transfer, treatments and
202
mode of delivery. Maternity wards were classified as type 3 when associated with a neonatal
203
intensive care unit (NICU). Steroids treatment was considered when the mother received at
204
least 1 injection of betamethasone.
205
Perinatal and 2-year outcomes 206
Perinatal outcomes included vital status, classified as TOP, antepartum stillbirth, death during
207
labor or in the delivery room (after spontaneous preterm labor or induction of labor), death in
208
the NICU9 and survival at discharge. We also investigated survival at discharge without
209
severe morbidity (i.e., without grade 3-4 intraventricular haemorrhage,10 cystic periventricular
210
leukomalacia,11 stage II or III necrotizing enterocolitis,12 stage 3 or greater retinopathy of
211
prematurity13 and/or laser treatment and severe bronchopulmonary dysplasia defined as
9
requiring oxygen for at least 28 days in addition to the requirement of 30% or more oxygen
213
and/or mechanical ventilator support or continuous positive airway pressure at 36 weeks’
214
postmenstrual age14). Z-score birth weights were calculated from EPOPé intrauterine growth
215
curves corrected for sex and gestational age.15 The third outcome was survival at 2 years’
216
corrected age without cerebral palsy whatever the stage. Cerebral palsy was defined according
217
to the diagnostic criteria of the Surveillance of Cerebral Palsy in Europe (SCPE) network.16
218
We thought to report deafness and blindness as well but there were no cases in our
219
population.8
220
Statistical analysis 221
We first compared characteristics and outcomes by type of pregnancy (single or multiple) and
222
found no significant difference, especially concerning median GA at PPROM, latency and
223
GA at birth, except for tocolysis and spontaneous onset of labor, which were significantly
224
more frequent in twins (Tables A.1 and A.2). Thereafter we analyzed singletons and twins
225
together. We described natural history of PPROM, antenatal management and perinatal
226
outcomes overall, then compared them by week of gestational age at PPROM. Data are
227
reported as percentages with 95% confidence intervals (95% CI) or medians with interquartile
228
range (IQR). Medians of quantitative variables were compared by a nonparametric
equality-229
of-medians test. When comparing by week of gestational age, to account for the
non-230
independence of twins, we used generalized estimating equations (GEE) to obtain p-values,
231
assuming an exchangeable correlation structure.17 To account for the duration of the
232
recruitment periods by gestational age at birth, a weighted coefficient was allocated to each
233
individual (1 for births at 22-26 weeks, 1.346 for births at 27-31 weeks and 7 for births at
32-234
34 weeks). Attrition is a key issue in longitudinal cohort studies.8 In this analysis, the
235
proportion of infants eligible but lost to follow-up was 17.7% of infants alive at 2 years’
236
corrected age (8.2% of all fetuses included). We compared characteristics of eligible infants
10
with and without follow-up and found no difference, except for low maternal age and low
238
socio-economic status that were associated with loss to follow-up (Table A.3). In addition to
239
complete-cases analysis, we performed multiple imputations with chained equations with a
240
logistic regression imputation model for missing binary data and a multinomial imputation
241
model for missing categorical data. Imputation model variables included both those
242
potentially predicting non-response and/or outcomes (type of maternity unit, maternal age and
243
country of birth, socioeconomic status, parity, gestational ages at PPROM and at birth,
244
latency duration, multiple pregnancy, in utero transfer, antenatal steroids and antibiotics,
245
magnesium sulfate, tocolysis, clinical chorioamnionitis, cord prolapse, placental abruption,
246
small for gestational age, cesarean section, sex, severe neonatal morbidities) and outcomes
247
(survival, cerebral palsy). Outcomes were estimated within each of the 30 imputed datasets
248
generated with 20 iterations, and results were pooled for a final analysis according to Rubin’s
249
rules. Statistical significance was set at two-tailed p < .05. Data were analyzed by use of
250
Stata/SE 13.0 (StataCorp LP, College Station, TX, USA).
251 252
Results
253
Among the 1435 women with a diagnosis of PPROM, 379 were at 22 to 25 weeks’ gestation,
254
with 427 fetuses alive (331 singletons and 96 twins) (Figure 1). Pregnancy was complicated
255
by PPROM at 22, 23, 24and 25weeks’ gestation in 101 (21.4%), 95 (24.1%), 99 (24.0%) and
256
132 fetuses (30.5%), respectively.
257
The overall population was 78% French or European, with a median age of 29 years
258
(interquartile range [IQR] 26-34), 91% lived with a partner and 51% were nulliparous, with
259
no significant difference by GA at PPROM (Table A.4).
11
Median GA at PPROM was 24 (IQR 23-25) weeks. Latency duration ranged from 0.5 to 145
261
days. Latency duration did not differ by week of GA at PPROM, nor did latency exceeding 2
262
days, 7 days or 14 days (Table 1). Whatever the GA at PPROM, nearly half of the fetuses
263
were born within the first week of latency. Consequently, GA at birth significantly increased
264
with GA at PPROM (Table 1). Only 5 infants (weighted percentage 7.1%) were born at 32-34
265
weeks. The overall weighted rates of placental abruption, cord prolapse and clinical
266
chorioamnionitis were 4.3% (95% CI 2.8-6.8), 2.9% (1.7-4.9) and 9.5% (7.0-12.8),
267
respectively. Eight fetuses (1.7% [0.9-3.4]) presented pulmonary hypoplasia and/or limb
268
deformities. The frequency of these complications did not differ by week of GA at PPROM.
269
Severe oligohydramnios was diagnosed in 217 fetuses (61.1% [55.3-66.7]), with increased
270
frequency for the earliest PPROM (61%, 76%, 57%, 53% at 22, 23, 24 and 25 weeks,
271
respectively, p=.05).
272
We found major differences in the obstetric management by GA at PPROM (Table 1). More
273
than 95% of infants were born in a type 3 maternity unit with PPROM at 24 or 25 weeks
274
versus 58% and 78% with PPROM at 22 and 23 weeks. Accordingly, rates of in utero transfer
275
were two- to threefold higher after 24 weeks. Most fetuses were exposed to antenatal steroids
276
and caesarean section when PPROM occurred after the threshold considered for neonatal
277
resuscitation in France in 2011 (24 weeks). The use of antenatal antibiotics, mainly
278
amoxicillin and 3rd generation cephalosporins, was lower at 22 weeks (81% vs > 92%
279
afterwards). Causes and indications for delivery were mainly spontaneous onset of labor
280
(62.2%) and induction of labor or cesarean section for clinical chorioamnionitis (18.5%).
281
With PPROM at 22-25 weeks, pregnancy outcomes were TOP (10 fetuses, 2.0%), antepartum
282
stillbirth (21 fetuses, 5.6%), death during labor (81 fetuses, 16.6%), death in the delivery
283
room (58 fetuses, 12.0%), death in the NICU (56 infants, 12.1%) or discharge alive (201
284
infants, 51.7%), with significant differences by GA at PPROM (Figure 1, Table 2). TOPs
12
were mostly performed for the earliest cases of PPROM (7, 1, 2 and 0 TOPs with PPROM at
286
22, 23, 24 and 25 weeks, respectively) complicated by anhydramnios and/or chorioamnionitis.
287
Stillbirths and deaths in the delivery room were mainly related to specific complications of
288
PPROM (clinical chorioamnionitis, oligohydramnios, placental abruption or cord prolapse) or
289
spontaneous delivery before 24 weeks. Deaths in the NICU occurred within the first week for
290
41% and within the first month for 84% of deceased children. These deaths were mostly
291
related to respiratory failure (38%), central nervous system injury (23%) or infection (14%).
292
Among the 315 liveborn infants, 68.2% survived until discharge, 51.6% survived until
293
discharge without severe morbidity (38.8% of all fetuses) and 58.9% were survivors at 2
294
years’ corrected age without cerebral palsy (43.4% of all fetuses). Overall, 13 infants had
295
cerebral palsy (1, 1, 7 and 4 with PPROM at 22, 23, 24 and 25 weeks, respectively) but none
296
had visual or auditory impairment. When considering all fetuses or liveborn infants, rates of
297
survival, survival at discharge without severe morbidity and survival at 2 years’ corrected age
298
without cerebral palsy significantly improved with increased GA at PPROM (Tables 2 and 3).
299
For example, among all fetuses, rates of survival at discharge were 14.1%, 39.5%, 66.8% and
300
75.8% with PPROM at 22, 23, 24 and 25 weeks, respectively. However, when focusing on
301
survivors at discharge or survivors at 2 years CA, survival at discharge without severe
302
morbidity or survival at 2 years’ corrected age without cerebral palsy did not differ by GA at
303
PPROM (Tables 2 and 3).
304 305 Comment 306 Main findings 307
This descriptive study shows that with PPROM at 22-25 weeks’ gestation, overall and for
308
each GA at PPROM, nearly half of the fetuses were delivered within the first week. Obstetric
309
management appears to be strongly influenced by GA at PPROM and by the threshold of
13
viability considered in France in 2011 (24 weeks’ gestation). Overall, PPROM at 22-25 weeks
311
was associated with high frequencies of perinatal mortality and morbidity. Both perinatal and
312
childhood prognosis, related to all fetuses or to liveborn infants, significantly improved with
313
advancing GA at PPROM: survival without cerebral palsy was low with PPROM at 22 and 23
314
weeks, but not zero, and reached approximately 60% and 70% with PPROM at 24 and 25
315
weeks. Nevertheless, incidences of severe morbidity and subsequent cerebral palsy by GA at
316
PPROM were similar among survivors, and potentially related to GA at birth and to postnatal
317
management taking GA at birth into consideration.
318
Strengths and limitations 319
The strengths of our study include a large sample of singletons and twins born preterm after
320
PPROM at 22-25 weeks, which allowed for reporting characteristics and outcomes stratified
321
by week of GA at PPROM, and follow-up at 2 years’ corrected age. Because singletons and
322
twins have similar latency durations and outcomes, our findings are relevant for both types of
323
pregnancies, even though the prognosis could slightly differ between twins with intact or
324
ruptured membranes. Unlike all published studies,2,4,18–20 our sample stems from a prospective
325
population-based cohort at a national level, thereby reflecting the diversity of antenatal
326
management and outcomes in “real-life” practices. Moreover, accounting for all pregnancy
327
outcomes when estimating neonatal prognosis allows for providing realistic figures that do
328
not overestimate the chances of survival. The use of different inception points and thus
329
denominators to report rates of survival is helpful in adapting information provided to parents
330
during pregnancy when the GA of birth is not yet known.21 Finally, the use of standardized
331
definitions for outcomes allows for comparison with other international studies or cohorts.21
332
The main limitation of this study is the proportion of missing data related to loss to follow-up
333
at 2 years’ corrected age, although attrition was moderate in relation to the cohort size and its
14
geographical extent.8 Appropriate statistical methods, with multiple imputations, allowed for
335
accounting for missing data and obtaining non-biased estimators. Another limitation, due to
336
the design of the EPIPAGE 2 cohort, involves left truncation and right-censoring of the
337
sample at 346/7 weeks.22 We avoided left truncation by including women with both PPROM
338
and delivery from 22 weeks. Concerning right-censoring, we likely missed the cases of
339
PPROM at 22-25 weeks for fetuses delivered at 35 weeks and afterwards. We assume that
340
such cases are exceptional and have a favorable neonatal prognosis. Their non-inclusion leads
341
to a very slight underestimation of the chances of survival or disease-free survival. A
342
disadvantage of these population-based data is that we are limited in investigating precisely
343
the medical teams’ willingness to provide antenatal active care (such as antenatal steroids or
344
performing a cesarean section), which can change as the pregnancy progresses. Moreover,
345
some specific complications, namely pulmonary hypoplasia, are likely underdiagnosed as
346
autopsies were not systematically performed to determine the cause of fetal or neonatal death.
347
Interpretation 348
Because of the high risks of extreme prematurity and severe disability, a key point in
349
antenatal care is to adequately inform parents facing PPROM at 22-25 weeks and to consider
350
their wishes in all decisions.1,3,5,23,24 However, in this context, the information given to parents
351
and the resulting management decisions depend very little on individual socioeconomic and
352
clinical characteristics (except for GA) but are largely influenced by the institution and the
353
practitioner who gives the information.24–28 There is indeed great variability in how caregivers
354
understand the prognosis of early PPROM, including neurodevelopmental impairment, and
355
their willingness to propose active management.26 This variability can be explained by
356
significant variations in published rates of survival with early PPROM, leaving practitioners
357
with a great uncertainty.
15
Indeed, reported survival after early PPROM ranges from 20% to 85%, survival without
359
severe morbidity from 20% to 70% and cerebral palsy from 0% to 10%.2,4,6,18–20 Many
360
reasons account for these variations. Selection bias, related to exclusion of women electing
361
TOP or immediate induction of labor as well as women not eligible for expectant
362
management or related to preadmission bias in tertiary-care referral centers, leads to
363
overestimating latency durations and survival rates.2,4,6,18–20 Ranges of GA at PPROM are
364
wide and differ widely across studies; hence, overall non-stratified results do not allow for
365
appropriate comparisons. Small sample sizes do not provide precise estimations.2,6,20 Finally,
366
published studies feature a retrospective design over 5 to 15 years,6,18,20 but medical practices
367
may have evolved and mortality rates may decrease.29 Therefore, comparing our findings with
368
previous publications is challenging.21
369
We report high rates of mortality and morbidity when preterm births occur following early
370
PPROM. Most children will be delivered extremely preterm, and their immaturity and
371
fragility are major risk factors of adverse outcomes. The frequency of the other obstetric
372
complications (placental abruption, cord prolapsed and chorioamnionitis) is lower than or
373
similar to that previously described.2,6,19,20 With PPROM at 22-25 weeks’ gestation, perinatal
374
outcomes appear to be influenced by medical practices, which are themselves affected by the
375
resuscitation threshold considered in France in 2011 (24 weeks).24,28,30,31 This hypothesis
376
requires further investigation.
377
Because French guidelines about management of women with PPROM are broadly similar to
378
those of other countries, our results may be generalizable to most developed countries with
379
similar practices and are relevant to question the strategies of management of early pregnancy
380
complications.32 Improving the prognosis of these pregnancies probably requires a rethinking
381
of care policies in a multidisciplinary way, involving obstetricians, neonatologists, care
382
networks, parent associations and policy makers.
16
Conclusion
384
Following PPROM, both parents and professionals are left with a great deal of uncertainty
385
regarding the evolution of pregnancy, complications and fetal and neonatal prognosis. Our
386
findings on the prognosis of PPROM at 22-25 weeks, based on prospective, population-based
387
data at a national level, provide new insights that can be used as a support for counseling
388
parents, especially during pregnancy when the GA of birth is not yet known. The impact of
389
the practitioner's decisions on the prognosis should lead to homogenize and optimize the
390
antenatal management practices.
391 392 393 394 395 396 397 398 399 400 401 402 403
17
Contribution to authorship:
404
Study concept: EL, GK
405
Study supervision: GK
406
Design of the present study: EL, GK
407
Acquisition, analysis and interpretation of data: All authors
408
Statistical analysis: EL, GK
409
Drafting of the manuscript: EL, GK
410
Critical revision of the manuscript for important intellectual content: All authors
411
Final approval of the version to be published: All authors
412
Agreement to be accountable for all aspects of the work: All authors
413
Acknowledgements: We are grateful to the participating children and their families and to all
414
maternity and neonatal units in France. The authors thank Laura Smales for editorial
415
assistance and acknowledge the collaborators of the EPIPAGE2 Obstetric writing group:
416
Pierre-Yves Ancel, MD, PhD (Inserm UMR 1153, Obstetrical, Perinatal and Pediatric
417
Epidemiology Research Team (Epopé), Center for Epidemiology and Statistics Sorbonne
418
Paris Cité, DHU Risks in pregnancy, Paris Descartes University, Unité de Recherche Clinique
419
– Centre d’Investigations Cliniques P1419, Département Hospitalo-Universitaire Risks in
420
Pregnancy, Cochin Hotel-Dieu Hospital, Assistance Publique-Hôpitaux de Paris, Paris
F-421
75014, France), Catherine Arnaud, MD, PhD (Research Unit on Perinatal Epidemiology,
422
Childhood Disabilities and Adolescent Health, INSERM UMR 1027, Paul Sabatier
423
University, Toulouse, France), Julie Blanc, MD (Department of Obstetrics and Gynecology,
424
Aix Marseille University, Marseille, France), Pascal Boileau, MD, PhD (Department of
425
Neonatal Pediatrics, Poissy Saint Germain Hospital, France, EA7285 Versailles Saint Quentin
426
en Yvelines University, France), Thierry Debillon, MD, PhD (Department of Neonatal
427
Pediatrics, University Hospital, Grenoble, France), Pierre Delorme, MD, MSc (Inserm UMR
428
1153, Obstetrical, Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for
429
Epidemiology and Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes
430
University, Department of Obstetrics and Gynecology, Cochin, Broca, Hôtel Dieu Hospital,
431
AP-HP, Paris, France), Claude D’Ercole, MD (Department of Obstetrics and Gynecology,
432
Nord Hospital, Assistance Publique des Hôpitaux de Marseille (AP-HM), Aix Marseille
433
Université, AMU, Marseille, France), Thomas Desplanches, RM, MSc (Inserm UMR 1153,
434
Obstetrical, Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for
435
Epidemiology and Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes
436
University), Caroline Diguisto, MD, MSc (Inserm UMR 1153, Obstetrical, Perinatal and
437
Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and Statistics
438
Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University, Maternité Olympe
439
de Gouges, University Francois Rabelais, Tours, France), Laurence Foix-L’Hélias, MD,
440
PhD (Inserm UMR 1153, Obstetrical, Perinatal and Pediatric Epidemiology Research Team
441
(Epopé), Center for Epidemiology and Statistics Sorbonne Paris Cité, DHU Risks in
442
pregnancy, Paris Descartes University, Sorbonne Universités, UPMC Univ Paris 06, IFD, 4
18
Place Jussieu, 75252 PARIS cedex 05, Paris, France, Department of Neonatal Pediatrics,
444
Trousseau Hospital, AP-HP, Paris, France), Aurélie Garbi, MD (Department of
445
Neonatology, Assistance Publique Hopitaux de Marseille, Marseille, France), Géraldine
446
Gascoin, MD, PhD (Department of Neonatal Medicine, Angers University Hospital, Angers,
447
France), Adrien Gaudineau, MD (Department of Obstetrics and Gynecology, Hautepierre
448
Hospital, Strasbourg, France), Catherine Gire, MD (Department of Neonatology, North
449
Hospital, Marseille, France), François Goffinet, MD, PhD (Inserm UMR 1153, Obstetrical,
450
Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and
451
Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University,
452
Department of Obstetrics and Gynecology, Cochin, Broca, Hôtel Dieu Hospital, AP-HP,
453
Paris, France), Gilles Kayem, MD, PhD (Inserm UMR 1153, Obstetrical, Perinatal and
454
Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and Statistics
455
Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University, Sorbonne
456
Universités, UPMC Univ Paris 06, IFD, 4 Place Jussieu, 75252 PARIS cedex 05, Paris,
457
France, Department of Obstetrics and Gynecology, Trousseau Hospital, AP-HP, Paris,
458
France), Bruno Langer, MD (Department of Obstetrics and Gynecology, Hautepierre
459
Hospital, Strasbourg, France), Mathilde Letouzey, MD, MSc (Department of Neonatal
460
Pediatrics, Poissy Saint Germain Hospital, France), Elsa Lorthe, RM, PhD (Inserm UMR
461
1153, Obstetrical, Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for
462
Epidemiology and Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes
463
University), Emeline Maisonneuve, MD, MSc (Department of Obstetrics and Gynecology,
464
Trousseau Hospital, APHP, Paris, France), Stéphane Marret, MD, PhD (Department of
465
Neonatal Medicine, Rouen University Hospital and Région-INSERM (ERI 28), Normandy
466
University, Rouen, France), Isabelle Monier, RM, PhD (Inserm UMR 1153, Obstetrical,
467
Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and
468
Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University), Andrei
469
Morgan, MD, PhD (Inserm UMR 1153, Obstetrical, Perinatal and Pediatric Epidemiology
470
Research Team (Epopé), Center for Epidemiology and Statistics Sorbonne Paris Cité, DHU
471
Risks in pregnancy, Paris Descartes University), Jean-Christophe Rozé, MD, PhD
472
(Department of Neonatal Medicine, Nantes University Hospital, Nantes, France,
473
Epidémiologie Clinique, Centre d'Investigation Clinique (CIC004), Nantes University
474
Hospital, Nantes, France), Thomas Schmitz, MD, PhD (Inserm UMR 1153, Obstetrical,
475
Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and
476
Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University,
477
Department of Obstetrics and Gynecology, Robert Debré Hospital, Assistance
Publique-478
Hôpitaux de Paris, Paris, France), Loïc Sentilhes, MD, PhD (Department of Obstetrics and
479
Gynecology, Bordeaux University Hospital, Bordeaux, France), Damien Subtil, MD, PhD
480
(Department of Obstetrics and Gynecology, Jeanne de Flandre Hospital, Lille, France),
481
Héloïse Torchin, MD, MSc (Inserm UMR 1153, Obstetrical, Perinatal and Pediatric
482
Epidemiology Research Team (Epopé), Center for Epidemiology and Statistics Sorbonne
483
Paris Cité, DHU Risks in pregnancy, Paris Descartes University, Neonatal Medicine and
484
Resuscitation Service in Port-Royal, Cochin Hospital, Assistance Publique-Hôpitaux de Paris,
485
Paris, France), Barthélémy Tosello, MD (Department of Neonatology, Assistance Publique
486
Hopitaux de Marseille, Marseille, France), Christophe Vayssière, MD, PhD (Department of
487
Obstetrics and Gynecology, University Hospital, Toulouse, France, Research Unit on
488
Perinatal Epidemiology, Childhood Disabilities and Adolescent Health, INSERM UMR 1027,
489
Paul Sabatier University, Toulouse, France), Norbert Winer, MD, PhD (Department of
490
Obstetrics and Gynecology, University Hospital, INRA, UMR 1280 Physiologie des
491
adaptations nutritionnelles,Nantes, France), Jennifer Zeitlin (Inserm UMR 1153, Obstetrical,
19
Perinatal and Pediatric Epidemiology Research Team (Epopé), Center for Epidemiology and
493
Statistics Sorbonne Paris Cité, DHU Risks in pregnancy, Paris Descartes University).
494 495
All the collaborators of the EPIPAGE2 Obstetric writing group have no conflict of interest or
496
compensation in relation with this article to disclose. All of them consented to such
497 acknowledgment. 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
20
References
516
1. Mercer BM. Preterm premature rupture of the membranes. Obstet Gynecol. 2003
517
Jan;101(1):178–93.
518
2. Waters TP, Mercer BM. The management of preterm premature rupture of the
519
membranes near the limit of fetal viability. Am J Obstet Gynecol. 2009 Sep;201(3):230–40.
520
3. ACOG. Practice Bulletin No. 172: Premature Rupture of Membranes. Obstet Gynecol.
521
2016 Oct;128(4):e165-177.
522
4. Dewan H, Morris JM. A systematic review of pregnancy outcome following preterm
523
premature rupture of membranes at a previable gestational age. Aust N Z J Obstet Gynaecol.
524
2001 Nov 1;41(4):389–94.
525
5. CNGOF. Recommandations pour la pratique clinique RPM 1999 [Internet]. [cited
526
2014 Jun 17]. Available from: http://www.cngof.asso.fr/D_PAGES/PURPC_06.HTM
527
6. Manuck TA, Eller AG, Esplin MS, Stoddard GJ, Varner MW, Silver RM. Outcomes
528
of Expectantly Managed Preterm Premature Rupture of Membranes Occurring Before 24
529
Weeks of Gestation: Obstet Gynecol. 2009 Jul;114(1):29–37.
530
7. Ancel P-Y, Goffinet F, EPIPAGE 2 Writing Group. EPIPAGE 2: a preterm birth
531
cohort in France in 2011. BMC Pediatr. 2014;14:97.
532
8. Pierrat V, Marchand-Martin L, Arnaud C, Kaminski M, Resche-Rigon M, Lebeaux C,
533
et al. Neurodevelopmental outcome at 2 years for preterm children born at 22 to 34 weeks’
534
gestation in France in 2011: EPIPAGE-2 cohort study. The BMJ [Internet]. 2017 Aug 16;358.
535
Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5558213/
536
9. Patel RM, Kandefer S, Walsh MC, Bell EF, Carlo WA, Laptook AR, et al. Causes and
537
Timing of Death in Extremely Premature Infants from 2000 through 2011. N Engl J Med.
538
2015 Jan 22;372(4):331–40.
539
10. Papile L-A, Burstein J, Burstein R, Koffler H. Incidence and evolution of
540
subependymal and intraventricular hemorrhage: A study of infants with birth weights less
541
than 1,500 gm. J Pediatr. 1978 Apr 1;92(4):529–34.
542
11. Volpe JJ. Brain injury in premature infants: a complex amalgam of destructive and
543
developmental disturbances. Lancet Neurol. 2009 Jan;8(1):110–24.
544
12. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al. Neonatal
545
necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg. 1978
546
Jan;187(1):1–7.
547
13. Classification of Retinopathy of Prematurity. THe international classification of
548
retinopathy of prematurity revisited. Arch Ophthalmol. 2005 Jul 1;123(7):991–9.
549
14. Jobe AH, Bancalari E. Bronchopulmonary Dysplasia. Am J Respir Crit Care Med.
550
2001 Jun 1;163(7):1723–9.
21
15. Ego A, Prunet C, Lebreton E, Blondel B, Kaminski M, Goffinet F, et al. [Customized
552
and non-customized French intrauterine growth curves. I - Methodology]. J Gynecol Obstet
553
Biol Reprod (Paris). 2016 Feb;45(2):155–64.
554
16. Cans C. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy
555
surveys and registers. Dev Med Child Neurol. 2000 Dec 1;42(12):816–24.
556
17. Ananth CV, Platt RW, Savitz DA. Regression Models for Clustered Binary
557
Responses: Implications of Ignoring the Intracluster Correlation in an Analysis of Perinatal
558
Mortality in Twin Gestations. Ann Epidemiol. 2005 Apr 1;15(4):293–301.
559
18. van der Heyden JL, van der Ham DP, van Kuijk S, Notten KJB, Janssen T, Nijhuis JG,
560
et al. Outcome of pregnancies with preterm prelabor rupture of membranes before 27 weeks’
561
gestation: a retrospective cohort study. Eur J Obstet Gynecol Reprod Biol. 2013
562
Sep;170(1):125–30.
563
19. Manuck TA, Varner MW. Neonatal and early childhood outcomes following early vs
564
later preterm premature rupture of membranes. Am J Obstet Gynecol. 2014
565
Sep;211(3):308.e1-6.
566
20. Kibel M, Asztalos E, Barrett J, Dunn MS, Tward C, Pittini A, et al. Outcomes of
567
Pregnancies Complicated by Preterm Premature Rupture of Membranes Between 20 and 24
568
Weeks of Gestation: Obstet Gynecol. 2016 Aug;128(2):313–20.
569
21. Rysavy MA, Marlow N, Doyle LW, Tyson JE, Serenius F, Iams JD, et al. Reporting
570
Outcomes of Extremely Preterm Births. Pediatrics. 2016 Sep 1;138(3):e20160689.
571
22. Lorthe E, Ancel P-Y, Torchin H, Kaminski M, Langer B, Subtil D, et al. Impact of
572
Latency Duration on the Prognosis of Preterm Infants after Preterm Premature Rupture of
573
Membranes at 24 to 32 Weeks’ Gestation: A National Population-Based Cohort Study. J
574
Pediatr. 2017 Mar;182:47-52.e2.
575
23. Kaempf JW, Tomlinson MW, Campbell B, Ferguson L, Stewart VT. Counseling
576
Pregnant Women Who May Deliver Extremely Premature Infants: Medical Care Guidelines,
577
Family Choices, and Neonatal Outcomes. Pediatrics. 2009 Jun 1;123(6):1509–15.
578
24. ACOG Committee on Practice Bulletins-Obstetrics. Obstetric Care Consensus No. 4:
579
Periviable Birth. Obstet Gynecol. 2016 Jun;127(6):e157–69.
580
25. Tucker Edmonds B, Krasny S, Srinivas S, Shea J. Obstetric decision-making and
581
counseling at the limits of viability. Am J Obstet Gynecol. 2012 Mar;206(3):248.e1-248.e5.
582
26. Edmonds BT, McKenzie F, Panoch J, Frankel RM. Comparing Neonatal Morbidity
583
and Mortality Estimates across Specialty in Periviable Counseling. J Matern-Fetal Neonatal
584
Med Off J Eur Assoc Perinat Med Fed Asia Ocean Perinat Soc Int Soc Perinat Obstet. 2015
585
Dec;28(18):2145.
586
27. McKenzie F, Robinson BK, Tucker Edmonds B. Do maternal characteristics influence
587
maternal–fetal medicine physicians’ willingness to intervene when managing periviable
588
deliveries? J Perinatol. 2016 Jul;36(7):522–8.
22
28. Diguisto C, Goffinet F, Lorthe E, Kayem G, Roze J-C, Boileau P, et al. Providing
590
active antenatal care depends on the place of birth for extremely preterm births: the EPIPAGE
591
2 cohort study. Arch Dis Child - Fetal Neonatal Ed. 2017 Jun 30;fetalneonatal-2016-312322.
592
29. Younge N, Goldstein RF, Bann CM, Hintz SR, Patel RM, Smith PB, et al. Survival
593
and Neurodevelopmental Outcomes among Periviable Infants. N Engl J Med. 2017
594
16;376(7):617–28.
595
30. Janvier A, Lantos J. Delivery room practices for extremely preterm infants: the harms
596
of the gestational age label. Arch Dis Child - Fetal Neonatal Ed. 2016 Sep 1;101(5):F375–6.
597
31. Rysavy MA, Li L, Bell EF, Das A, Hintz SR, Stoll BJ, et al. Between-Hospital
598
Variation in Treatment and Outcomes in Extremely Preterm Infants. N Engl J Med. 2015 May
599
7;372(19):1801–11.
600
32. Smith LK, Blondel B, Reempts PV, Draper ES, Manktelow BN, Barros H, et al.
601
Variability in the management and outcomes of extremely preterm births across five
602
European countries: a population-based cohort study. Arch Dis Child - Fetal Neonatal Ed.
603 2017 Feb 23;fetalneonatal-2016-312100. 604 605 606 607 608 609 610 611 612 613 614 615 616 617
23
List of tables:
618
Table 1: Obstetric and neonatal characteristics by gestational age (GA) at PPROM
619
Table 2: Outcomes by GA at PPROM
620
Table 3: Outcomes at 2 years corrected age by GA at PPROM
621
Table A.1: Comparison of characteristics between singleton and twin pregnancies 622
Table A.2: Comparison of neonatal characteristics and outcomes between singleton and twin 623
pregnancies 624
Table A.3: Comparison of infants with and without follow-up at 2 years
625
Table A.4: Maternal characteristics by GA at PPROM
626 627 628 629 630 631 632 633
24 GA at PPROM Total 22 w 23 w 24 w 25 w p-value n (%) n (%) n (%) n (%) n (%) Characteristics N=427 N=101 N=95 N=99 N=132 Obstetric characteristics
GA at birth (w) median (IQR) (n=427) 25 (24-27) 23 (22-24) 24 (24-28) 25 (24-27) 26 (26-28) <.001 GA at birth among survivors at discharge (w)
median (IQR) (n=201) 27 (26-29) 28 (26-29) 28 (26-32) 27 (25-29) 26 (26-28) .17 GA at birth (w) (n=427) 22-23 95 (19.4) 67 (64.1) 28 (23.8) - - <.001 24-26 235 (48.1) 24 (23.0) 50 (42.4) 78 (66.4) 83 (55.7) 27-29 74 (20.4) 8 (10.3) 11 (12.6) 16 (18.3) 39 (35.2) 30-34 23 (12.1) 2 (2.6) 6 (21.2) 5 (15.3) 10 (9.1)
Latency (d) median (IQR) (n=427) 8.0 (2.9-20.9) 6.1 (2.4-16.0) 9.0 (2.4-31.0) 8.0 (3.2-21.0) 8.3 (2.9-19.0) .82
Latency > 2d (n=427) 332 (80.6) 77 (77.0) 69 (77.9) 78 (82.1) 108 (83.9) .57
Latency > 7d (n=427) 197 (53.0) 45 (46.4) 43 (55.9) 44 (53.2) 65 (55.0) .62
Latency > 14d (n=427) 121 (36.7) 26 (28.2) 30 (44.8) 26 (37.9) 39 (35.2) .31
Obstetric management
Born in type 3 maternity unit (n=427) 348 (83.8) 57 (57.9) 69 (77.9) 94 (95.8) 128 (97.3) <.001 Antenatal discussion of care limitation
(n=422) 97 (21.6) 38 (37.1) 23 (25.4) 22 (18.9) 14 (9.8) <.001 In utero transfer (n=425) 207 (49.8) 21 (21.3) 33 (34.6) 67 (71.0) 86 (64.9) <.001 Antibiotics (n=424) 394 (93.5) 81 (81.3) 86 (92.3) 98 (100.0) 129 (98.0) - Tocolysis (n=424) 246 (57.7) 27 (26.8) 46 (41.8) 71 (75.7) 102 (77.5) <.001 Corticosteroids (n=424) 274 (68.7) 26 (28.2) 44 (56.3) 84 (88.8) 120 (91.3) <.001 Magnesium Sulfate (n=418) 13 (3.1) 2 (2.6) 1 (0.9) 3 (2.9) 7 (5.2) .34 Spontaneous labor (n=426) 277 (62.6) 69 (68.0) 70 (71.9) 65 (57.6) 73 (55.5) .13 Caesarean delivery (n=423) 154 (39.2) 11 (12.5) 21 (22.3) 41 (49.6) 81 (62.7) <.001 Cephalic presentation (n=395) 218 (56.0) 43 (51.9) 45 (53.1) 54 (58.2) 76 (58.9) .74 Neonatal characteristics Male (n=424) 238 (56.9) 60 (61.6) 45 (45.7) 56 (60.8) 77 (59.4) .24
Birth weight (g) median (IQR) (n=409) 799 (630-1043) 560 (500-730) 730 (630-1120) 795 (680-1060) 900 (780-1090) <.001 Birth weight < 10th percentile (n=408) 72 (19.3) 14 (15.0) 10 (10.3) 17 (25.9) 31 (23.6) .049
GA: gestational age, PPROM: preterm premature rupture of membranes, w: weeks’ gestation, IQR: interquartile range, SD: standard deviation, d: days 635
Data are n (%) unless indicated. Percentages are weighted by recruitment period. 636
25 GA at PPROM
Total 22 w 23 w 24 w 25 w p-value
Outcomes n/N (%) [95%CI] n/N (%) [95%CI] n/N (%) [95%CI] n/N (%) [95%CI] n/N (%) [95%CI] Perinatal death among all fetuses
Termination of pregnancy 10/427 (2.0) [1.1-3.8] 7/101 (6.7) [3.2-13.4] 1/95 (0.9) [0.1-5.9] 2/99 (1.7) [0.4-6.6] 0/132 <.001 Antepartum stillbirth 21/427 (5.6) [3.1-9.8] 9/101 (8.6) [4.5-15.8] 4/95 (8.5) [2.2-28.2] 4/99 (3.4) [1.3-8.9] 4/132 (2.9) [1.1-7.6] Death during labor or in delivery
room 139/427 (28.6) [24.4-33.2] 65/101 (62.6) [52.5-71.6] 49/95 (41.6) [30.3-53.8] 16/99 (13.6) [8.3-21.6] 9/132 (6.3) [3.3-11.7] Death in NICU 56/427 (12.1) [9.3-15.5] 8/101 (8.0) [4.0-15.3] 11/95 (9.6) [5.2-17.1] 17/99 (14.5) [8.9-22.7] 20/132 (15.1) [9.9-22.3] Survival at discharge
Among all fetuses 201/427 (51.7) [46.3-57.1] 12/101 (14.1) [8.2-23.3] 30/95 (39.5) [26.8-53.7] 60/99 (66.8) [56.1-76.1] 99/132 (75.8) [67.7-82.3] <.001 Among liveborn infants 201/315 (68.2)
[62.6-73.4] 12/44 (31.1) [18.8-46.9] 30/58 (62.1) [46.9-75.3] 60/88 (73.7) [63.1-82.2] 99/125 (79.7) [71.7-85.9] <.001 Survival at discharge without severe morbidity*
Among all fetuses 140/418 (38.8) [33.3-44.7] 9/101 (10.6) [5.6-19.2] 19/94 (29.5) [17.4-45.4] 36/95 (46.8) [34.5-59.6] 76/128 (60.6) [51.8-68.8] <.001 Among liveborn infants 140/306 (51.6)
[45.2-58.0] 9/44 (23.3) [12.7-39.0] 19/57 (46.7) [30.1-64.1] 36/84 (51.9) [38.8-64.7] 76/121 (63.9) [54.8-72.0] <.001 Among survivors at discharge 140/192 (76.7)
[69.9-82.3] 9/12 (75.0) [44.2-91.9] 19/29 (75.7) [56.0-88.5] 36/56 (71.5) [57.2-82.5] 76/95 (80.8) [71.6-87.6] .68 GA: gestational age, NICU: neonatal intensive care unit, PPROM: preterm premature rupture of membranes, w: weeks’ gestation
639
All percentages obtained with complete-cases analysis, denominators can vary slightly accordingly to missing data, namely for survival at discharge without 640
severe morbidity (9 missing data). 641
* Survival at discharge without severe morbidity is defined as survival at discharge without grades 3-4 intraventricular haemorrhage, cystic periventricular 642
leukomalacia, stages II or III necrotizing enterocolitis, stage 3 or greater retinopathy of prematurity and/or laser treatment and severe bronchopulmonary 643
dysplasia. 644
26 GA at PPROM
Outcomes Total 22 w 23 w 24 w 25 w p-value
% (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI)
Death after discharge (n=201) 1.2 (0.4-3.7) 0 0 1.3 (0.2-8.7) 1.8 (0.4-6.9) - Cerebral palsy among survivors at 2-years’ corrected age
CC (n=163) 7.2 (4.1-12.3) 11.2 (1.5-50.4) 3.2 (0.4-20.5) 11.8 (5.4-24.1) 5.0 (1.8-12.7) .41 MI (n=198) 9.1 (4.5-13.7) 13.1 (0.0-35.4) 5.8 (0.0-14.7) 13.1 (4.0-22.3) 7.1 (0.9-13.2) .62 Survival at 2-years’ corrected age without cerebral palsy
Among all fetuses
CC (n=392) 43.4 (37.6-49.4) 10.5 (5.6-19.1) 36.0 (23.2-51.1) 55.5 (43.2-67.2) 66.3 (57.0-74.5) <.001 MI (n=427) 46.4 (40.8-52.1) 12.3 (5.2-19.4) 37.2 (23.2-51.1) 57.3 (45.8-68.8) 69.1 (60.8-77.5) <.001 Among liveborn infants
CC (n=280) 58.9 (52.4-65.1) 24.0 (13.0-40.0) 57.9 (41.5-72.7) 61.8 (49.0-73.1) 70.4 (60.9-78.4) <.001 MI (n=315) 61.3 (55.2-67.3) 27.1 (12.9-41.2) 58.5 (43.0-74.0) 63.2 (51.7-74.8) 72.7 (64.4-81.0) <.001 Among survivors at 2 years’ corrected age
CC (n=163) 92.8 (87.7-95.9) 88.9 (49.6-98.5) 96.8 (79.5-99.6) 88.2 (75.9-94.6) 95.1 (87.3-98.2) .41 MI (n=198) 90.9 (86.3-95.5) 86.9 (64.6-100.0) 94.2 (85.3-100.0) 86.9 (77.7-96.0) 92.9 (86.8-99.1) .62 GA: gestational age, PPROM: preterm premature rupture of membranes, w: weeks’ gestation, CC: complete cases analysis, MI: multiple imputation 647
Missing data for cerebral palsy at 2-years’ corrected age are related to 3/201 deaths after discharge, and 35/198 children lost to follow-up. Percentages of 648
cerebral palsy and survival without cerebral palsy were obtained using multiple imputations for missing data. 649 650 651 652 653 654 655 656 657
27 Table A.1: Comparison of characteristics between singleton and twin pregnancies
659
Singletons Twins p-value
N=331 N=96
Maternal characteristics
Maternal age (y) median (IQR) (n=426) 29 (26-34) 29 (26-32) .99 Born in France/Europe (n=406) 243 (78.3) 70 (78.6) .97
Marital life (n=413) 287 (90.3) 88 (95.4) .29
Tobacco use (n=412) 89 (27.5) 16 (17.4) .16
Nulliparous (n=426) 150 (47.6) 60 (62.7) .06
Obstetric characteristics
GA at PPROM (w) median (IQR) (n=427) 24 (23-25) 24 (23-25) .77 GA at birth (w) median (IQR) (n=427) 25 (24-28) 25 (24-27) .80 GA at birth among survivors at discharge (w)
median (IQR) (n=201)
27 (26-30) 27 (25-28) .66 Latency (d) median (IQR) (n=427) 8.0 (2.8-23.0) 8.0 (2.9-18.0) .91
Latency > 2d (n=427) 256 (80.4) 76 (81.1) .88
Latency > 7d (n=427) 153 (53.5) 44 (50.8) .65
Latency > 14d (n=427) 89 (36.6) 32 (38.1) .82
Obstetric management
Born in type 3 maternity (n=427) 266 (83.0) 82 (86.8) .50 Antenatal discussion of care limitation (n=422) 81 (23.4) 16 (15.1) .20
In utero transfer (n=425) 155 (48.7) 52 (53.8) .52 Antibiotics (n=424) 302 (92.8) 92 (96.2) .37 Tocolysis (n=424) 174 (52.6) 72 (76.0) .004 Corticosteroids (n=424) 210 (68.6) 64 (69.1) .95 Magnesium Sulfate (n=418) 13 (3.9) 0 (0) - Spontaneous labor (n=426) 197 (57.2) 80 (82.2) .003 Cesarean delivery (n=423) 111 (36.6) 43 (48.5) .13 Cephalic presentation (n=395) 168 (56.1) 50 (55.5) .92
GA: gestational age, PPROM: preterm premature rupture of membranes, w: weeks’ gestation, IQR: interquartile range, SD: standard deviation, d: days, y: 660
years 661
Data are n (%) unless indicated. Percentages are weighted by recruitment period. 662
28 Table A.2: Comparison of neonatal characteristics and outcomes between singleton and twin pregnancies
665
Singletons First twin Second twin p-value
N=331 N=48 N=48
Neonatal characteristics
Male (n=424) 187 (57.2) 23 (51.7) 28 (60.0) .56
Birth weight (g) Median (IQR) (n=409)
800 (635-1060) 730 (580-1000) 800 (620-1030) .76 Birth weight < 10th percentile
(n=408)
51 (18.1) 11 (24.9) 10 (22.6) .59
Perinatal death among all fetuses
Termination of pregnancy 8 (2.1) 1 (1.9) 1 (1.9) .74
Antepartum stillbirth 17 (6.0) 3 (6.3) 1 (1.9)
Death during labor or in delivery room
116 (30.4) 12 (22.7) 11 (20.8)
Death in NICU 42 (11.5) 6 (12.0) 8 (16.5)
Survival at discharge
Among all fetuses (n=427) 148 (50.0) 26 (57.1) 27 (58.9) .51 Among liveborn infants (n=315) 148 (66.9) 26 (74.5) 27 (71.1) .65 Survival at discharge without severe morbidity
Among all fetuses (n=418) 112 (40.7) 14 (31.9) 14 (32.6) .46 Among liveborn infants (n=306) 112 (54.8) 14 (41.9) 14 (39.5) .17 Among survivors at discharge
(n=192)
112 (83.1) 14 (57.0) 14 (56.3) .002
Survival at 2-years’ corrected age without cerebral palsy
Among all fetuses (n=392) 104 (40.3) 22 (53.2) 24 (55.4) .17 Among liveborn infants (n=280) 104 (55.7) 22 (71.4) 24 (67.3) .21 Among survivors at 2 years old
(n=163)
104 (89.2) 22 (100.0) 24 (96.6)
-GA: gestational age, NICU: neonatal intensive care unit, PPROM: preterm premature rupture of membranes, w: weeks’ gestation 666
All percentages obtained with complete-cases analysis, denominators can vary slightly accordingly to missing data, namely for survival at discharge without 667
severe morbidity (9 missing data) and survival at 2-years’ corrected age without cerebral palsy (35 missing data). 668
* Survival at discharge without severe morbidity is defined as survival at discharge without grades 3-4 intraventricular haemorrhage, cystic periventricular 669
leukomalacia, stages II or III necrotizing enterocolitis, stage 3 or greater retinopathy of prematurity and/or laser treatment and severe bronchopulmonary 670
dysplasia. 671
29 Cerebral palsy data available
among survivors at 2-years CA eligible for the study
Yes (n=163) No (n=35) p-value
Characteristics n (%) n (%)
Maternal characteristics
Maternal age (n=198) median (IQR) 29 (26-33) 27 (22-30) .006 Born in France/Europe (n=194) 120 (76.7) 22 (70.7) .53
Parents’ socio-economic status (n=189)* <.001
Professional 36 (25.7) 1 (2.9)
Intermediate 27 (15.3) 0 (0)
Administrative, public service, self-employed, students
51 (31.4) 10 (34.4) Shop assistants, service workers 25 (13.5) 3 (9.8)
Manual workers 17 (12.5) 16 (52.9) No known occupation 3 (1.6) 0 (0) Nulliparous (n=197) 84 (54.0) 13 (37.0) .10 Obstetric characteristics GA at PPROM (w) (n=198) 22 10 (5.8) 2 (6.8) .33 23 26 (20.1) 4 (10.9) 24 50 (32.3) 9 (24.3) 25 77 (41.8) 20 (58.0) GA at birth (w) (n=198) 22-23 0 (0) 0 (0) .81 24-26 93 (44.3) 21 (52.7) 27-29 55 (35.3) 8 (27.0) 30-34 15 (20.4) 6 (20.3)
Latency (d) median (IQR) (n=198) 17.5 (6.0-31.2) 17.2 (4.0-23.0) .79
Twin pregnancy (n=198) 47 (26.2) 6 (15.9) .39
Placental abruption (n=198) 11 (5.9) 2 (6.8) .91
Cord prolapse (n=198) 5 (2.6) 1 (2.5) .90
Obstetric management
Born in type 3 maternity unit (n=198) 161 (99.1) 35 (100.0) .54
In utero transfer (n=198) 105 (64.4) 22 (60.4) .52
30 Tocolysis (n=198) 116 (68.9) 24 (67.2) .97 Corticosteroids (n=198) 151 (93.5) 32 (92.5) .72 Magnesium Sulfate (n=196) 7 (3.9) 2 (6.9) .49 Caesarean delivery (n=196) 99 (62.3) 18 (51.3) .36 Neonatal characteristics Male (n=198) 93 (59.5) 20 (58.0) .95
Birth weight < 10th percentile (n=198) 29 (21.5) 8 (23.6) .83
Severe bronchopulmonary dysplasia (n=182) 23 (13.1) 6 (18.8) .30 Severe necrotizing enterocolitis (n=195) 5 (2.9) 1 (2.6) .71 Severe retinopathy of prematurity (n=198) 6 (2.9) 2 (5.9) .55 Severe cerebral lesion (IVH and/or cPVL)
(n=198)
14 (7.0) 2 (5.0) .71
* Defined as the highest occupational status of the mother and father, or mother only if living alone. 673
CA: corrected age, cPVL: cystic periventricular leucomalacia, GA: gestational age, IVH: intraventricular hemorrhage, PPROM: preterm premature rupture of 674
membranes, w: weeks’ gestation, d: days, IQR: interquartile range 675
Data are n (%) unless indicated. Percentages are weighted by recruitment period. 676 677 678 679 680 681 682 683 684 685 686 687 688